Burr regression and portfolio segmentation

被引：26

作者：

Beirlant, J

Goegebeur, Y

Verlaak, R

Vynckier, P

机构：

[1] Katholieke Univ Leuven, Dept Math, B-3000 Louvain, Belgium

[2] Katholieke Univ Leuven, Dept Appl Econ, B-3000 Louvain, Belgium

[3] Boels & Begault Re AON Grp, B-1140 Brussels, Belgium

来源：

INSURANCE MATHEMATICS & ECONOMICS | 1998年 / 23卷 / 03期

关键词：

Burr distribution; regression; risk tarification;

D O I：

10.1016/S0167-6687(98)00045-6

中图分类号：

F [经济];

学科分类号：

02 ;

摘要：

Two Burr regression models are proposed. These models extend existing log-logistic regression models. An algorithm for computing the maximum likelihood estimators is proposed. Graphical techniques for model validation are incorporated. An actuarial application to portfolio segmentation for fire insurance is included. (C) 1998 Elsevier Science B.V. All rights reserved.

引用

页码：231 / 250

页数：20

共 12 条

[1] Tail index estimation, pareto quantile plots, and regression diagnostics [J].